ALGIRDAS GREIMAS seeks in his writing to find the "deep structure" of all narrativity. As a result, he is less interested in what Roland Barthes terms the proairetic and hermeneutic codes than he is in the formal elements in a narrative that create implicit (if not always consciously recognized) oppositions. To tie Greimas to Barthes once again, Greimas could be said to explore those codes that Barthes sees as outside of the mere forward temporal progression of narrative; that is, the symbolic, semantic, and cultural codes. (See the Barthes module on the 5 codes.) Greimas' influences in understanding the deep structure of narrative include Vladimir Propp's exploration of the deep structure of folklore, Claude Lévi-Strauss' work on the structure of myth, and Etienne Souriau's work on theater. Like the structural linguists that also inspire him, Greimas looks for what we might call the underlying grammar of narrative, "the semiolinguistic nature of the categories used in setting up these [narratological] models" (63); he wishes to find behind any "manifestation of narrativity" a "fundamental semantics and grammar" (65). Greimas is also interested in extending the relevance of narratology to all experience: "Our own concern... has been to extend as much as possible the area of application of the analysis of narrative" (63). Greimas therefore can be found applying his narratological models to phenomena that we might think fall outside of structural rules, for example passion (as he does in "On Anger: A Lexical Semantic Study" or The Semiotics of Passions). Greimas can do so because of a foundational precept of post-Saussurian linguistics: all language is arbitrary. There is no connection (other than convention) that links linguistic signs like writing or speaking to their referents. The sounds or written lines that make up the word "cat" have only an arbitrary, conventional connection to the actual cat that exists in the world. "Because of this," Greimas writes, "linguists became aware of the possibilities of a generalized semiotic theory that could account for all the forms and manifestations of signification" (17). Anything that we as humans articulate in language (which is to say, pretty much everything) should therefore conform to structural rules: in this principle, we find the heart of Greimas' discipline, semiotics or the study of signification (which is to say, the study of the use of signs to refer to things). To put this another way, the connection between signification and the real world is completely arbitrary; however, signification is in itself not arbitrary since language tends to follow structural rules. Humans are therefore caught in a system of rules and deep structures that bear no relation to the real world. This disjunction between language and reality is, in fact, a central precept of contemporary theory, from Jacques Lacan's understanding of the real to Judith Butler's understanding of performativity to Jean Baudrillard's theorization of the simulacrum. Lacan, Butler, and Baudrillard all do rather disparate things with the knowledge of language's arbitrary relation to the real world. Greimas' goal is purely structuralist: he wishes to find the deep structures by which all signification orders the world of perception. As Greimas puts it, "From this perspective the sensible world as a whole becomes the object of the quest for signification. As long as it takes on form, the world appears, as a whole and in its various articulations, as potential meaning. Signification can be concealed behind all sensible phenomena; it is present behind sounds, but also behind images, odors, and flavors, without being in sounds or in images (as perceptions)" (17).

When it comes to analyzing specific narratives (say, a novel), Greimas is interested in seeing how any specific instance of narrativity relates to a larger process of general meaning-making (or semiosis). A good example is the distinction he makes (following Propp) between actors and actants. An actor refers to the actual character that appears in a narrative; however, such "actors" tend to follow the structural logic of narrative deep structures. They always fulfill the position of structural functions and thus are related to "actants." A given character can, for example, serve the function of "acting subject" or "passive object." A narrative action can be said to imply a character who is the "sender"; a character who serves as the "object"; and a character that serves as "receiver." Building on Propp's Morphology of the Folktale, Greimas also looks at the actantial opposition, helper/opponent; that is, those characters that serve either to aid or hinder the quest of the protagonist. Consider, for example, the love triangle, a pervasive narrative convention in which we find two male characters competing with each other in order to gain the love of a woman, who serves as the passive object of their competition. Were the two male characters to enter into a fight because of the woman, the two characters would change places as senders and receivers (of punches, invective, etc.). On a structural level, the woman serves merely as the passive object of their contending affections. (Eve Sedgwick makes much of this structural relation in her understanding of homosociality.) As this example illustrates, a single character can at different times in the narrative serve different actantial roles: "one actant can be manifested by several actors and, conversely, one actor can at the same time represent several actants" (111), as Greimas puts it. We can understand this relation of actor to actant as a spectrum of possibilities: "If we polarize these observations we can theoretically conceive of two extreme types of possible actorial structures: (a) Actorial manifestation can have a maximal expansion characterized by there being an independent actor for every actant or actantial role (a mask, for example, is an actor having the modality of seeming as its actantial role). We would say that, in this case, the actorial structure is objectivized. (b) Actorial distribution can show a minimal expansion and be reduced to just one actor responsible for all of the necessary actants and actantial roles (giving rise to absolute interior dramatization). In this case, the actorial structure is subjectivized" (112-13). A good example of the former case is religious allegory (e.g., Spenser's Faerie Queene), which tends to reduce characters to types. A good example of the latter extreme is James Joyce's Finnegan's Wake, which occurs completely inside the unconscious of a single "actor."

The situation with regard to actants is made even more complex because of the tendency of any one actant to suggest or give rise to its opposite: you can have a positive subject in a narrative as well as a "negative subject" or "anti-subject"; a positive object (for example, the good woman) can be opposed to a negative object (for example, the tempting femme fatale). In a given narrative, value judgments may be applied to these oppositions (hero vs. traitor, for example); however, what is important, according to Greimas, is the structural relation, so that the same oppositions will occur in purely aesthetic productions where value-judgments are largely withheld. Any number of "modalities" can also be applied to actants: for example, the actant's degree of competence ("the wanting and/or being-able and/or knowing-how-to-do of the subject" [109]) or the degree of simulation ("The overdetermination of actants according to this category of being and seeming accounts for the extraordinary game of disguises (jeu de masques), which includes confrontations between heroes who might be hidden, unrecognized, or recognized and disguised traitors who are unmasked and punished" [111]). Strider in The Lord of the Rings is a good example of both of these modalities, since he questions his competence as king early in the tale and is, for most of the first book, an ambiguous character (his identity is quasi-secret).

According to Greimas, any narrative is merely a manifestation of such deep structures: "narrative forms are no more than particular organizations of the semiotic form of the content for which the theory of narration attempts to account" (114). As a result, he claims that his structural principles apply both to the most complex narratives consisting of thousands of pages and to the most minimalist of narrative units. Even a single word entails a limited panoply of related terms that could potentially be strung out across a narrative: "Thus, to take a familiar example, the figure sun organizes around itself a figural field that includes rays, light, heat, air, transparency, opacity, clouds, etc." (115). (To illustrate this point, I have attempted, under "Applications," to apply Greimassian principles to the simple sentence, "The road is clear,") Because of this relation between the structural oppositions implied in characters and the thematic oppositions implied by any given word, Greimas argues that we can posit a structural relation between a given narrative's characters (narrative structures) and themes (discursive structures): "A character in a novel, supposing that it is introduced by the attribution of a name conferred on it, is progressively created by consecutive figurative notations extending throughout the length of the text, and it does not exist as a complete figure until the last page, thanks to the cumulative memorizing of the reader" (119). Peter Brooks' theory of narrative is strongly influenced by this sort of semiotic principle. (See the Brooks module on plotting.) For Greimas, "An actor is... a meeting point and locus of conjunction for narrative structures and discursive structures, for the grammatical and the semantic components" (120).

What links thematic oppositions and actantial oppositions is the fact that any theme or actant automatically entails its opposite, thus creating a field of potential oppositions linked together in what Greimas calls the "semiotic square." For more, see the next module on the semiotic square.